Google Fights Porn With Copyright-ID Tech

The fight against child pornography is getting an assist from technology designed by Google Inc. to help identify copyright-protected clips on its YouTube video-sharing site.

Four Google employees used their "20 percent time" - during which the company encourages them to pursue unofficial, out-of-the-box projects - to customize the copyright software for the National Center for Missing and Exploited Children's program for identifying children in sexually explicit photos and video.

Software already has been used to match known images, but when new ones are submitted by law enforcement and service providers, analysts generally make identifications manually, often based on recollections. The center has recently employed facial-recognition technology to make those IDs easier, but with limited success.

With the new Google tools, analysts can also seek matches based on other attributes, such as the color and shape of a couch or the wallpaper pattern in the background of a photo.

Ernie Allen, the center's president, appreciates the new tool.

The center used to be "dependent on the memory of the analyst who was already overwhelmed by the sheer number of content we were receiving," he said, adding that its 11 analysts processed 5 million child-pornography images last year.

Allen said the center's work helps prosecutors build evidence and police locate children who have been forced to engage in sexual acts.

Google's contribution grew out of an existing partnership between the center and leading technology companies, including Yahoo Inc., Time Warner Inc.'s AOL and Microsoft Corp.

"The keys here were organization, scalability and search," Google research scientist Shumeet Baluja wrote about the project on a company blog. "In particular, the tools we provided will aid in organizing and indexing NCMEC's information so that analysts can both deal with new images and videos more efficiently and also reference historical material more effectively."